Autoimmune Encephalitis: a Disease of the 21St Century at the Crossroads of Neurology and Psychiatry V

ISSN 1819-7124, Neurochemical Journal, 2018, Vol. 12, No. 4, pp. 311–323. © Pleiades Publishing, Ltd., 2018. Original Russian Text © V.V. Fominykh, E.A. Frei, L.V. Brylev, N.V. Gulyaeva, 2018, published in Neirokhimiya, 2018, Vol. 35, No. 4, pp. 308–322.

REVIEW ARTICLES

Autoimmune Encephalitis: A Disease of the 21st Century at the Crossroads of Neurology and Psychiatry V. V. Fominykha, b, 1, E. A. Freic, L. V. Bryleva, b, d, and N. V. Gulyaevaa, d aInstitute of Higher Nervous activity and Neurophysiology, Russian Academy of Sciences, Moscow, Russia bBuyanov City Clinical Hospital, Moscow Department of Healthcare, Moscow, Russia cOslo University Hospital, Oslo, Norway dResearch and Clinical Centre for Neuropsychiatry, Moscow Department of Healthcare, Moscow, Russia Received May 3, 2018

Abstract⎯Autoimmune encephalitis is a group of neurological diseases characterized by brain damage by autoantibodies towards extra- or intracellular structures of the nervous system that act as antigens. The combination of neurological and mental disorders, as well as the ability to identify a specific “antigen and antibody” axis, makes these diseases extremely interesting from the standpoint of “molecular psychiatry” and the creation of new experimental models of cognitive processes, clinical diagnosis and targeted treatment. How- ever, despite active research in this direction and a large number of specific antibodies, the diagnosis of autoimmune encephalitis is often extremely difficult and can be made only by clinical criteria. This study provides an overview of the available data on the history of discovery and study of autoimmune encephalitis, describes the methods for searching of antibodies that exist today and the further prospects for studying this group of diseases.

Keywords: autoimmune encephalitis, NMDA, history, autoimmune diseases of the nervous system DOI: 10.1134/S1819712418040037

“The brain is a monstrous, beautiful mess.” chiatry,” when a discovery of a specific substrate explained the development of mental disorders. Susannah Cahalan, Brain on Fire: Following the first discovery, many laboratories My Month of Madness started to work in this area, and, currently, one or two new antibodies are described every year [2]. The occurrence of each of these antibodies may lead to the INTRODUCTION occurrence of psychiatric symptoms in patients: acute One of the most exciting discoveries in neurology psychosis, as well as behavioral and cognitive impair- of the 21st century is the isolation of antibodies to ments. n-methyl-D-aspartate receptors (NMDAR) and the In approximately 60% of patients, psychiatric identification of the cause of acute psychosis and symptoms appear at the beginning of the disease and behavioral disorders in young patients with ovarian become leading in clinical picture [3]. Usually these teratoma [1]. The isolation of these antibodies is one patients are hospitalized or treated by psychiatrists, of the small bridges in the big world of “molecular psy- often with the diagnosis of “schizophrenia,” while correct diagnosis and immunosuppressive therapy 1 Corresponding author; address: ul. Butlerova, 5A, Moscow, may completely cure the patient and return them to a 117485 Russia; phone: +7(495)3347020; e-mail: hydrohi- normal life. Dramatic effect of therapy and the cur- [email protected]. Abbreviations: AIT, autoimmune thyroiditis; AE, autoimmune ability of these states lead to isolation of a separate encephalitis; MRI, magnetic resonance imaging; CSF, cerebro- group of autoimmune encephalitis and the active spinal fluid; AMPA, α-amino-3-hydroxy-5-methyl- 4-isoxaz- study of pathogenic antibodies and substrates. ole propionic acid; AMPAR, AMPA receptor; CASP2R, contactin-associated protein 2 receptor; DPPX, dipeptidyl pepti- Autoimmune encephalitis (AE) is a group of neu- dase-like protein; D2R, dopamine receptor 2; GABAaR, rological diseases characterized by brain damage by GABAa receptor; GABAb, GABAb receptor; HEK, human autoantibodies to the extra- or intracellular structures embyronic kidney cells; LGI1, Leucine-rich, glioma inactivated of the nervous system, which act as antigens [4, 5]. 1 protein; mGlu5R, metabotropic glutamate receptor 5; NMDA, N-methyl-D-aspartate; NMDAR, NMDA receptor; This group of diseases is divided into two large sub- PANDAS, pediatric autoimmune neuropsychiatric diseases due groups: paraneoplastic encephalitis, whose develop- to streptococcal infection. ment is associated with the presence of a tumor and

311 312 FOMINYKH et al.

Table 1. Classical “paraneoplastic” encephalitides with antibodies against intracellular antigens Antigen Clinical symptoms (most often) Tumor References Anti-Hu (ANNA-1) Encephalitis (limbic, cortical, brain- >90% (small-cell lung cancer) 33 stem), polyneuropathy, autonomic impairments MA (MA1, MA2) Encephalitis (limbic, brainstem), poly- >90% (ovary, breast, colon cancers) 55 neuropathy Amphiphysin Limbic encephalitis, encephalopathy, >90% (small-cell lung cancer, mammary 36 stiff-person syndrome, degeneration of adenocarcinoma) the cerebellum Anti-Ri (ANNA-2) Limbic encephalitis, encephalopathy, >90% (small-cell lung cancer) 35 stiff-person syndrome, degeneration of the cerebellum, dystonia CV2 (CRMP5) Encephalitis, optic neuritis, retinitis, >90% (small-cell lung cancer, thymoma) 37 polyneuropathy, myelopathy, Lambert- Eaton syndrome, degeneration of the cerebellum, movement disorders (chorea and other) Anti-GAD antibodies Encephalitis (limbic, cortical, brain- <5% (thymoma, kidney cancer, 73 stem), stiff-person syndrome, degenera- mammary or colon adenocarcinoma) tion of the cerebellum Other paraneoplastic neurological syndromes ZIC-4 Subacute degeneration of the cerebellum Small-cell lung cancer 74 Yo-1 (PCA-1) Subacute degeneration of the cerebellum Often (ovary cancer, breast cancer) 34 Tr-receptor Subacute degeneration of the cerebellum Hodgkin’s lymphoma 73 SOX-1 Lambert-Eaton syndrome Small-cell lung cancer 74 subsequent antigen presentation, and idiopathic (1) antibodies to neurotransmitter receptors encephalitis, where autoantibodies are produced with- (NMDAR, AMPAR, mGluR5, GABAaR, GABAbR, out any association with any oncological process [4]. GlyR, and dopamine receptor 2 (D2R)); Three groups of antibodies were separated using (2) antibodies to transmembrane proteins the same principle: (CASPR2, DPPX, IgLON5, Neurexin3a, etc.); (1) antibodies to intracellular antigens or, as they (3) antibodies to secreted proteins (Leucine-rich, are often called, “classical” paraneoplastic antigens glioma inactivated 1 protein, and LGI1). (Hu, Yo, Ma2, CV2, and amphiphysin), which in According to the presence of antibodies and their most cases are associated with the presence of a type, it is possible to predict the rate of progression of tumor; the disease, the prognosis, and the response to ther- (2) antibodies to the surface neuronal membrane apy, which makes a search for specific antibodies that may be detected in both paraneoplastic and idio- extremely important in each case. pathic cases (metabotropic glutamate receptor 5 (mGlu5R), GABAb receptor (GABAbR), NMDAR, In addition, the discovery of antibodies to surface α-amino-3-hydroxy-5-methyl- 4-isoxazole propionic neuronal antigens led to a burst of new ideas in the acid (AMPAR), and contactin-associated protein 2 field of psychiatry: the hope arose that the molecular receptor (CASP2R); and cellular bases of psychiatric diseases may be unraveled and the theory of the relationship between (3) antibodies that are more often detected idio- immunological disorders and the development of pathically (glutamate decarboxylase (GAD65), dipep- schizophrenia received a new area of development [7]. tidyl peptidase-like protein (DPPX), IGloN5, glycine In addition to the pathogenetic role of antibodies, lev- receptor (GlyR)). els of AE antibodies in acute psychoses began to be Tables 1 and 2 show current list of antibodies and actively studied. It was revealed that in some cases possible diverse clinical manifestations. In addition, patients with acute psychosis have antibodies to for AEs with antibodies to the surface neuronal mem- NMDAR (from 1.46 to 20 % according to various brane, there is a different classification of antigen studies, depending on the criteria for selection of localization [6]: group of patients and methods for antibodies evalua-

NEUROCHEMICAL JOURNAL Vol. 12 No. 4 2018 AUTOIMMUNE ENCEPHALITIS 313

Table 2. Autoimmune encephalitides with antibodies to surface neuronal membrane or synaptic proteins Antigen Clinical symptoms Tumor References NMDAR Prodromal symptoms, psychiatric disturbances, epi- Ovary teratoma, rarely carcinoma 1 leptic seizures, memory impairments, paroxysmal (10–45% of all cases depending movement episodes, catatonia, autonomic disorders, on age) in HSV-encephalitits sur- consciousness decline, coma, hypoventilation vivors. Idiopathic AMPAR Limbic encephalitis, psychiatric disturbances 70% (pulmonary cancer, breast 43 cancer, thymoma) GABAbR Limbic encephalitis, often seizures 50% (pulmonary cancer, neuro- 44 endocrine cancer) GABAaR Refractory seizures, status epilepticus or Rarely, reported in thymoma 48 Kozhevnikov epilepsy, stiff-person syndrome, opso- clonus LGI1 Limbic encephalitis, 60% hyponatremia, faciobra- <10% (pulmonary cancer, thy- 41 chial dystonic seizures moma) CASPR2 Encephalitis, Morvan syndrome, neuromyotonia <40% (thymoma) 41 GlyR Stiff-person syndrome, progressing encephalomyeli- Rarely 42 tis with rigidity and myoclonus, limbic encephalitis, degeneration of the cerebellum, optic neuritis IgLON5 Motor impairments during sleep, behavioral distur- No 51 bances, obstructive sleep apnea, respiration disorders, dysarthria, dysphagia, ataxia, chorea DPPX Diarrhea, consciousness decline, psychiatric disor- No 47 ders, tremor, myoclonus, epileptic seizures, nystag- mus, hyperekplexia, ataxia, encephalomyelitis with rigidity and myoclonus mGluR5 Limbic encephalitis, Ophelia syndrome Often, Hodgkin’s lymphoma 45 D2R Encephalitis with involvement of basal ganglia, Rarely 16 Sydenham’s chorea Neurexin3 NMDA-like encephalitis, orofacial dyskinesia Reported idiopathically and after 52 malaria tion). Based on these studies, “red flags” were identi- syphilis doctors caught the idea that “madness and fied, which can help to diagnose AE in acute psycho- insanity” may have a biological cause: at the end of the ses, and the possibility of immunotherapy also started 19th century, patients with delirium, schizophreni- to be discussed more actively [3]. form disorders, and neurosyphilis were observed by all Although most antibodies have been described in psychiatrists in Europe [9]. Neurosyphilis and its psy- the last 15–20 years and AE is a pathology of the 20th chiatric manifestations, viz., dementia, depression, and 21st centuries, the first conceptions on the possi- delirium, psychosis, episodes of arousal, hallucina- ble organic nature of mental illnesses, and subse- tions, depersonalization, and cognitive impairments, quently of their autoimmune etiology, occurred much were the most frequent reason for hospitalization in earlier. The chronological sequence of the AE study is psychiatric clinics from the late 19th century to the shown in Fig. 1. 1940s until penicillin was discovered [10]. The frequency of psychiatric manifestations in neurosyphilis THE HISTORY OF THE DISCOVERY is extremely high, from 33 to 86% [11]; however, the OF AUTOIMMUNE ENCEPHALITIS mechanism of their development in some cases is not completely clear. Previously, it was confirmed by stud- The Era of Clinical Observations ies of pathological autopsy material of the brain that and Descriptions of Individual Nosologies there are lesions and Treponema pallidum is present in Although neurosyphilis is not an autoimmune dis- these regions. This led to the formation of views on the ease and is a well-known infectious process with a spe- relationship between persistent infection and the cific pathogen, Treponema pallidum, we start our development of manifestations of neurosyphilis [12]. review from it. During the treatment of patients with However, recent studies using magnetic resonance

NEUROCHEMICAL JOURNAL Vol. 12 No. 4 2018 314 FOMINYKH et al. IgLON5 DPPX 2010 mGLuR5 GABAb AMPA GLyR CASPR LGi1 CASPR NMDA VKCG 2000 MA CV2 Antibodies surface to neuronal membrane Ri 1990 amphiphysin Hu Yo 1980 Classical paraneoplastic antibodies 1970 1960 Hashimoto’s encephalopathy Hashimoto’s Limbic encephalitis 1950 1940 1930 The chronological sequence of the encephalitides. The study sequence of autoimmune chronological 1920 Fig. 1. 1910 von Economo’s von encephalitis lethargica encephalitis 1900 1890 Paraneoplastic encephalitis Paraneoplastic Sydenham’s chorea Sydenham’s 1880

NEUROCHEMICAL JOURNAL Vol. 12 No. 4 2018 AUTOIMMUNE ENCEPHALITIS 315 imaging (MRI) and other methods of neuroimaging (1) the pandemic of encephalitis after the “Spanish have shown that neurosyphilis, like AE, is often asso- flu” pandemic (influenza that affected approximately ciated with normal MRI or brain atrophy of varying 30% of the world’s population at the beginning of the severity, as well as a pronounced antibody response in 20th century); the form of immunoglobulins in the cerebrospinal (2) the relationship between pharyngitis in 55% of fluid (CSF) [9]. This suggests that psychotic disorders patients with the presence of elevated titers to anti- may develop not only in the presence of focal and bac- streptolysin; terial damage but also for other reasons: (3) the pathomorphological data of the brain stud- (1) the immediate toxic effect of Treponema on ies of the deceased from encephalitis lethargica, which neurons, which disturbs the normal functioning of showed the presence of inflammatory lymphocytic neurons and cortex; infiltrates in the midbrain and basal nuclei, as well as (2) production of antibodies due to a persistent the presence of oligoclonal antibodies in the CSF, the chronic infection, which interact with different presence of antibodies to the basal nucleus antigens, molecular targets and lead to the development of psy- and a positive response to steroid therapy [16]. chiatric disorders. In addition, immunohistochemical analysis of spo- Many studies have shown that the permeability of radic cases of encephalitis lethargica in later periods the blood–brain barrier is impaired in neurosyphilis showed the presence of antibodies mainly against neu- and antibodies to Treponema pallidum are synthesized rons, as well as T- and B-cell infiltration in the basal intrathecally [13], while the exact mechanism of the nucleus region. However, sometimes it cannot be occurrence of mental disorders in patients without ruled out that antibodies against the basal nuclei focal neurological pathology and lesion areas in MRI occurred after their damage by some other pathologi- is currently not known. In addition, according to cal process [17]. a number of studies with antibiotic therapy, the classi- Sydenham’s chorea. Sydenham’s chorea was cal picture of neurosyphilis and generalized paralysis, described by Paracelsus as “the dance of St. Witt,” which was previously well studied, changed [14], and whereas in the 17th century Sydenham gave it an accu- syphilis is still a “great imitator.” rate clinical description, and he also described acute Similarly, in a number of other infectious patholo- rheumatic fever. However, the relationship between gies, researchers have discussed the theory of the the two diseases was suggested later, in the 18th cen- development of schizophrenic-like mental disorders tury, by Richard Bright and further supported by a and their relationship to antibodies (for example, in number of researchers. Researchers at that time wrote: patients with antibodies to Toxoplasma gondii [15]). “Of two children with rheumatic fever, one is neces- sarily sick with chorea.” The classical description of Neurosyphilis was one of the first and most com- Sydenham’s chorea is a combination of chorea, some- mon diseases in which psychiatric manifestations was times with the development of paresis, and behavioral assumed to have a clear structural basis in the form of disorders (mainly, the obsessive-compulsive spec- a brain lesion caused by Treponema pallidum. The first trum) after an infection caused by group A hemolytic described AEs include encephalitis lethargica, Syden- streptococcus. ham’s chorea, and Hashimoto’s encephalopathy. In At the end of the 20th century, the expanded term addition, Rasmussen’s encephalitis and FIRES syn- PANDAS (pediatric autoimmune neuropsychiatric drome are historically considered within the frame- diseases due to streptococcal infection) was intro- work of AE-like diseases; however, due to almost duced. The proposed mechanism for the development complete lack of knowledge about the pathogenesis of of these diseases is the cross-reactivity of antibodies to these diseases and the availability of screening anti- streptococcus and basal ganglia, which leads to extra- bodies, these diseases will not be considered in this pyramidal movement disorders and psychiatric mani- review [6]. festations [16]. Recently, a number of studies consid- Encephalitis lethargica. Historically, von Economo’s ered Sydenham’s chorea as encephalitis with antibod- encephalitis lethargica, whose pandemia was observed ies to D2R [18]. from 1915 to 1925 after the epidemic of influenza, is Hashimoto’s Encephalopathy. In 1966, Brain et al. assumed to have postinfectious autoimmune genesis. described a case of episodic encephalopathy associ- Due to the sudden development and the same sudden ated with autoimmune thyroiditis (AIT). A few disappearance of the disease, its causes are virtually months after the detection of AIT, the patient (a 48-year unknown; however, clinically, patients developed a old man) periodically had pareses of different localiza- picture of weakness, fatigue, sleep disorders such as tions, a decreased level of consciousness, and cogni- lethargy, ophthalmoplegia, catatonia, and symptoms tive impairments that persisted despite therapy with of parkinsonism and psychosis. hormones, anticoagulants, and thyroxine; however, 1 Assumptions on the autoimmune genesis of year later it had regressed completely. Doctors sug- encephalitis lethargica were made on the basis of sev- gested the relationship between AIT and brain damage eral clinical and epidemiological facts: on the background of thyroid pathology, as the

NEUROCHEMICAL JOURNAL Vol. 12 No. 4 2018 316 FOMINYKH et al. described picture of the disease did not fit into any of Oncology Development and Description the known pathologies at that time [19]. of the Group of Paraneoplastic Encephalitis Further, many attempts have repeatedly been made The next era in the study of AE is a description of to classify these conditions, develop criteria, and the group of paraneoplastic encephalitides (Table 1). understand the etiology of Hashimoto’s encephalopa- Classification and studies of this group of diseases thy. However, due to the rarity of disease and began in the same era as the molecular description of unknown mechanism of brain damage, these investi- various types of tumors in the 1950s. Prominent gations were not successful. researchers in autoimmune neurology believe that the first case of autoimmune paraneoplastic encephalitis The main clinical manifestations include epileptic was described by Oppenheim in 1888 [25] in the article seizures, resistant to antiepileptic therapy, headache, “Neurological symptoms associated with carcinomatosis hallucinations, stroke-like episodes and other focal without detected changes in the brain.” A 54-year-old neurological symptoms, impairments of cognitive woman was first admitted to the Charite Clinic in Ber- functions and consciousness up to coma, behavioral lin in June 1887. Her neurological status included and mood disorders, ataxia, and dementia. The fre- agnosia, mood changes, and severe aphasia; commu- quency of occurrence of this disease is 2.1/100000 (the nication was possible only with the simplest gestures. National Institute of Health of the United States clas- A few days later, the patient died, autopsy revealed sifies this disease as rare); it occurs more frequently in gastric cancer; however, there were no visible changes women, has a fluctuating nature of pathology, and in the brain, including meningeal membranes and ves- responds to immunosuppressive therapy. The autoim- sels. Histological examination revealed no significant mune genesis of encephalopathy is also confirmed by changes such as metastases and vascular or nerve a number of laboratory tests. The patients have a high pathologies. Given the strange combination of neuro- titer of antibodies to thyroperoxidase or thyroglobulin, logical symptoms without any pathological changes, whereas in most cases euthyroidism or a slight change Oppenheim began to look for similar cases in the in thyroid function was observed [20]. A number of archives of the clinic and found a similar description studies showed the presence of antibodies in the CSF 1 year before where a woman with epileptic seizures which specifically bound to cerebellar astrocytes. The and breast cancer had no changes in a pathological CSF often had an elevated level of IgG and/or oligo- morphological study of brain tissue. Oppenheim sug- clonal bands [21]. Several studies have shown the pres- gested the presence of a “toxic effect” of the tumor on ence of immunoreactive receptors of thyroid-stimu- the brain, even in the absence of metastases. In 1929, lating hormone in the cerebral cortex [22]. In 2002, subacute cerebellar degeneration was described in antibodies to alpha-enolase were found [23], which patients with oncological pathology and only half a were detected in 68 to 83.3% patients with century later, in 1948, Denny-Brown [26] described a Hashimoto’s encephalopathy according to various rapidly progressing sensory neuronopathy in patients studies [24]. with lung cancer and suggested the need for “metabolic studies in these cases to identify the cause of However, the antineuronal autoimmune response lesions of the nervous system.” In 1949, the term is not fully confirmed: the data of single-photon emis- “paraneoplastic” was introduced by Guichard and sion computer tomography on brain hypoperfusion, Vignon to describe the neuropathy that arose in a high expression of alpha-enolase in the vascular endo- patient with uterine cancer [27]. thelium and the presence of antibodies to it in other vasculitis (systemic lupus erythematosus, rheumatoid After Oppenheim the second cases with damage to arthritis, etc.) served as basis for a hypothesis on the the central but not peripheral nervous system was vascular genesis of Hashimoto’s encephalopathy [21]. described by Brierley and his colleagues in 1960: three cases of the disease with an inflammatory response in Accordingly, currently, there are three views on the the cerebral hemispheres, limbic system and concom- mechanism of development of Hashimoto’s encepha- itant bronchial carcinoma, kidney leiomyoma, and an lopathy [21]: unknown lung tumor, respectively [28]. The same case is considered as the first description of the clinical pic- (1) autoimmune vasculitis of the central nervous ture of limbic encephalitis. Later, several similar cases system; were described; in 1968 Corsellis [29] introduced the (2) autoimmune encephalitis with antibodies term “limbic encephalitis” and wrote a review of the against thyroid and central nervous system antigens; literature on this topic. At the same time, three hypotheses for the devel- (3) metabolic pathology due to the toxic effect of opment of paraneoplastic encephalitis have been pro- thyrotropic hormone on the central nervous system. posed: However, nowadays, a number of researchers (1) neurodegeneration of cells for an unknown rea- doubt not only the genesis but also the existence of this son, which leads to secondary inflammatory infiltra- pathology. tion;

NEUROCHEMICAL JOURNAL Vol. 12 No. 4 2018 AUTOIMMUNE ENCEPHALITIS 317

(2) a possible viral infection; the NMDAR agonist in the central nervous system, (3) main hypothesis of autoimmune lesion of the D-serine [7]. However, the antibody–antigen–depen- nervous system by antibodies [30]. dent response that led to these symptoms in patients was described by Dalmau. At this stage, we note only the description of clinical cases and the study of the morphological pattern of First, a group of 12 women, who developed psychi- pathology in the brains of deceased people. Anti-Hu atric behavioral disorders, amnesia, dyskinesias, epi- antibodies were discovered only 100 years after the leptic seizures, autonomic disorders, as well as a clinical description of Oppenheim. decrease in the level of consciousness or even coma, and the need for ventilation support, were clinically In 1985, the first work was published on the detection of antibodies to various neuronal structures in described. After the laboratory examination, it was biological fluids: antibodies to Purkinje cells in para- found that the serum and CSF of patients were immunoreactive against neuropil of the hippocampus and neoplastic degeneration of the cerebellum were described [31] and neuronal antinuclear antibodies the forebrain; luminescence was the strongest at the associated with sensory neuropathy and lung cancer, level of membrane and the molecular layer of the hippocampus. Antibodies were isolated and found to subsequently described as anti-Hu antibodies [32, 33]. react with NR1/NR2 NMDAR heteromers. After fur- Several anti-neuronal antibodies have been ther examination, 11 patients had ovarian teratomas described: and 1 patient had mediastinal teratoma, all the exam- in 1983, anti-Yo antibodies [34]; in 1989, anti-Ri ined tumors expressed the NR2B subunit of NMDAR. antibodies [35] in patients with breast cancer; it was Tumor removal and immunotherapy resulted in com- also shown that isolated IgG bound to tumor cells. plete recovery in eight of nine patients; three patients In 1993, antibodies to amphiphysin (128 kDa brain without tumor removal died from neurological aggra- protein) were isolated in three patients with Stiff-per- vations. During autopsies, glial changes in brain tissue, son syndrome and breast cancer [36], and in 1996, rare T-cell infiltrates, and neuronal degeneration, anti-CV2 antibodies (antibodies to the 66 kDa pro- which was more pronounced in the hippocampus, tein) [37] and antibodies to Ma2 in patients with tes- were revealed. ticular tumor. In addition to the description of NMDA encepha- At this stage, a standard protocol for antibodies litis, in 2005, six patients with new antibodies to potas- searching has already been developed (see the Meth- sium channels were described, which in further studies ods section). divided into two subtypes: with antibodies against CASP2R and LGI1 [41], who had a good response to However, the description of these antibodies and therapy and removal of the tumor. the precise diagnosis of the associated neurological syndrome did not lead to effective treatment of these Several years after the description of the clinical patients. In most cases, therapy was ineffective and the picture and antibodies to NMDAR, it became clear prognosis was disappointing. This situation was that there are many cases of AE. Many of these cases changed only after description of first AE with anti- remain unrecognized due to the onset of the disease bodies to the surface neuronal antigens and good with psychiatric symptoms, a normal MRI picture or recovery after treatment. minor changes in MRI. Antibodies searching and discovery are very important due to treatable state of these conditions. NMDAR, CASP2R, LGI1 and Description of Encephalitides other new antibodies were isolated into a separate with Antibodies against Neuronal Surface group of antibodies to the surface neuronal mem- The next wave of encephalitis research began in brane. 2005, when the Dalmau et al. for the first time clini- In 2008, progressive encephalomyelitis with rigid- cally described paraneoplastic syndrome in young ity and myoclonus was associated with antibodies to women with ovarian teratomas and then isolated anti- GlyR [42]. It is interesting that in this case the clinical bodies to NMDAR [38, 39] (Table 2). picture of encephalitis with hyperekplexia, which was The idea of a relationship between NMDAR and described as a genetic disease caused by the presence schizophrenic manifestations had been already of a mutation in the glycine receptor gene, led to the known; it was found in the 1980s [40], when patients idea of the presence of antibodies to GlyR in this dis- could experience episodes of arousal, psychotic and ease, which was confirmed. cognitive impairments after the use of NMDAR In 2009, Lai et al. described antibodies to AMPAR antagonists (such as ketamine). Observation of the in limbic encephalitis [43], which were detected by effect of these drugs led to the creation of the theory of screening of CSF and serum of patients in neuronal hypofunctioning of NMDAR in schizophrenia, and culture by immunoprecipitation and immunoblotting. one of the animal models of schizophrenia was created In 1994, antibodies to the third subunit of the gluta- by the “switching off” of the gene that encodes D-ser- mate receptor subunit, which were detected in Ras- ine racemase, which is required for the production of mussen’s encephalitis and epilepsy with frequent sei-

NEUROCHEMICAL JOURNAL Vol. 12 No. 4 2018 318 FOMINYKH et al. zures, were described; however, subsequently the wide and based more on clinical picture, while the pathogenic role of these antibodies was not con- therapy is chosen empirically and often based on the firmed. existing expert opinion [53]. In 2010, in [44], during screening of 410 patients, 15 had an antibody response to the hippocampus; Identification and Characterization of Autoantibodies, however, no other known antibodies were detected. Confirmation of their Phatogenicity Serum and CSF of these patients were screened on rat hippocampal cultures and immunoreactivity was also From the description of Sydenham’s chorea and detected in relation to neuronal culture. Electrophore- until the first discovery of anti-Hu antibodies, AE has sis with immunoprecipitation revealed 90 and 105 kDa been described clinically. A macro and microscopic bands corresponding to GABAb1R and GABAb2R. picture of the brain pathology has also been studied, Subsequently, GABAbR was determined by mass and in some cases no significant changes have been spectroscopic analysis of immunoprecipitates. The found. technique described in this work in 2010 has become It’s interesting to note that names of the first anti- quite a standard scheme for searching for antibodies in bodies were simply combined from the first letters of AE with small modifications. patient’s names and surnames (Hu, Ri, etc.). It high- In 2011, antibodies to mGluR5 were detected in lights once again that clinical picture played the main Ophelia syndrome [45] and Hodgkin’s disease, role in a process of searching for new antibodies and although the syndrome itself was clinically described describing syndromes. in 1982 by the psychiatrist Carr in his teenage daughter The entire group of paraneoplastic antibodies was [46]. In this case, a technique analogous to the previ- isolated and studied using the standard scheme [35]: ous work on the detection of antibodies to GABAb2R (1) Screening of biological fluids (CSF and blood was used. serum) on frozen sections of the brain of deceased In 2013, Boronat et al. described antibodies to people without any neurological diseases, usually the DPPX [47]. zones of the neuropil of the hippocampus and the cor- In 2014, Mar Petit-Pedrol et al. described antibodies tex [33], or on brain tissue of mice or rats [54]. Some- to GABAaR also during screening work; in 6 patients times, the antigen could be predicted even at this out of 140 patients with suspected encephalitis, sei- stage, based on the characteristic localization of anti- zures or epileptic status and an autoimmune response to bodies on the membrane or binding of antibodies to a brain sections of unknown genesis, antibodies to GAB- particular brain region. This was the case, for example, AaR were detected during immunoprecipitation [48]. in the detection of antibodies to D2R after detection of However, prior to this, a similar picture of the dis- immunoreactivity against the basal nuclei. ease was suggested in two patients in 2012 on the basis (2) When detecting the immunoreactivity of indi- of the clinical picture; four mutations in GABAaR vidual sera or CSF, these samples, in parallel with the were previously described, which led to generalized control samples, were analyzed by immunoblotting. epilepsy [49]; this clinically also suggested the pres- At the same time, the reactivity of different dilutions of ence of antibodies against different variants of CSF and patients’ serum against protein extracts GABAR. obtained from homogenized brain tissue or from pre- In 2015, after the clinical analysis of sleep disorders isolated cortical neurons was evaluated. Thus, anti-Hu and PSG patterns, antibodies to IgLON5 were antibodies have been described to antigen 35–38 kDa. detected by the same method [50], the spectrum of At that time, a mass spectrometric determination clinical manifestations for which was subsequently of such amounts of antigen was technically difficult; expanded [51]. thus, the antigen was determined by immunoaffinity In 2016, the same group of researchers [52] chromatography. The isolated IgGs to the antigen described antibodies to neurexin 3; during a 10-year were applied to the column and a protein extract from study, five patients with a similar clinical picture were neurons of the cerebral cortex was passed through it, identified in the form of prodromal symptoms, head- which made it possible to obtain concentrated antigen ache, and symptoms of the gastrointestinal pathology, and confirm its mass on a western blot and conduct a and then development of epileptic seizures and lower- mass spectrometric analysis of the concentrated pro- ing of consciousness up to coma and the need for ven- tein [35]. tilation support. The picture of this encephalitis was In a number of cases (for example, the description described as NMDA-like, taking the prodromal symp- of antibodies to Ma2), modifications of the method toms and pronounced impairment of consciousness were applied: the antigen was determined using the into account. “phage display” method [55]. Currently, a further search for antibodies is con- Since 2010, when the determination of an antigen tinuing. The concepts of neurobiology of AE have has become available via mass spectrometry and the been created and may be used to study pathology at work of Lancaster et al. [44] was published, all further various levels [6]. However, diagnostic criteria are very studies were performed according to a standard proce-

NEUROCHEMICAL JOURNAL Vol. 12 No. 4 2018 AUTOIMMUNE ENCEPHALITIS 319 dure and mainly as screening. For only two antibodies, an antigen was predicted on the basis of clinical man- 1: Immunoreactivity of the CSF and blood ifestation (hyperekplexia in the case of encephalitis serum in brain sections or neuronal culture with antibodies to the glycine receptor) and a possible (immunohisto- and cytochemistry) common pathogenic antibody IgLON5 for a group of patients with a pattern of sleep disorders according to polysomnography. 2: Immunoprecipitation of antibodies The search for antibodies is usually performed and estimation of target antigens according to the standard scheme (Fig. 2). using immunoblotting The first stage is determination of immunoreactivity of patients’ serum or CSF using a neuronal culture 3: Mass-spectrometry of isolated bands and/or immunohistochemical staining of rodent brain with assay of an antigen sections. Immunocytochemical analysis is usually performed as follows: a rat hippocampal neuronal culture is incubated with antibodies to the NR1 subunit of NMDAR (control) or with CSF from patients. Then 4: Conﬁrmation of speciﬁcity of binding it’s possible to check if fluorescent signals are local- of isolated antibodies with the antigen: ized in the same areas [39]. The complex containing (a), study in cells transfected with the target the autoantibody is isolated from the cell culture and antigen; (b), study in wild-type the protein is identified by high performance liquid or target antigen knockout mice chromatography with tandem mass spectrometry. It’s important to remember that detected proteins may itself be complexes of antibodies with other proteins. Therefore, direct binding of antibodies is evaluated on 5: Study of isolated antibodies effects in vitro transfected cell cultures expressing candidate antigens. Unfortunately, it is not always possible to isolate 6: Study of isolated antibodies effects in vivo antigens from neuronal cultures, despite the rapid development of immunoproteomics. The biochemical properties of a number of proteins, such as insolubility in detergents (NMDAR), or post-transcriptional 7: Development of diagnostic assay modifications (pronounced lipidation or phosphory- lation) make it difficult to purify and analyze by mass Fig. 2. The standard protocol for the search, isolation, and spectrometry. In addition, some patient antibodies confirmation of pathogenicity of antibodies in autoim- bind weakly to rodent antigens-orthologues, which mune encephalitides. leads to false-negative results [6]. The second stage–immunohistochemical staining cultures, the methods for obtaining them have been of brain sections or cultures of rodent neurons with described in detail in the literature [56], or human antibodies isolated from patients. Unlike the first embryonic kidney cells line 293 (Human Embryonic method mentioned above, in this case two lines of ani- Kidney, HEK293). Cell lines transfected with plas- mals (or two types of cultures) are used. The first is the mids with the genes of the studied antigens, are useful control one, the second is the knockout for the gene of not only for evaluation of the direct binding of anti- interest (similarly for cell cultures: for example, the bodies, but also for characterization of their reactivity staining of the control line and the line with the against individual receptor subunits. Thus, Dalmau et knockdown of the gene of interest is compared). If a al., 2007 [38] used the HEK293 cell line to study the knockout or knockdown leads to the disappearance of reactivity of autoantibodies against NMDAR sub- fluorescence, this confirms that the patient has anti- units. The cells were transfected with plasmids con- bodies specific to the particular protein. taining the genes NR1, NR2A, or NR2B (alone or in However, the presence of autoantibodies in the combination), as well as a control plasmid without serum/CSF of the patient does not itself indicate their inserts. As a result, the cells expressed individual sub- pathogenicity. Therefore, the third mandatory stage of units of NMDAR (NR1, NR2A, or NR2B), as well as antibodies characterization is confirmation of their their functionally active combinations (NR1/NR2B pathogenicity. The simplest indirect confirmation of and NR1/NR2A). The resulting cells were incubated pathogenicity is the correlation between the antibody with serum or CSF of patients, followed by immunocy- titer and the severity of the symptoms. More reliable tochemistry. All 12 samples of patients’ serum and CSF evidences may be obtained in vitro and/or in vivo. A patients were reactive against NR1/NR2 heteromers common tool for in vitro analysis is cell cultures, that contained NR2B. Serum and CSF samples from which are used already at the stage of immunocyto- eight patients were also reactive against NR1/NR2A chemistry; these are usually rat or mouse neuronal heteromers. Serum samples were not reactive to cells

NEUROCHEMICAL JOURNAL Vol. 12 No. 4 2018 320 FOMINYKH et al. expressing individual receptor subunits. There was cells: for example, antibodies to GABAaR selectively also no reactivity against immunoblots of proteins reduce the amplitude and frequency of miniature from cells expressing functional receptors; that is, inhibitory postsynaptic potentials [57] and also reduce intact functional heteromers were required for anti- synaptic currents mediated by GABAR [6]. In another body binding. All these data suggested that the main experiment on cerebellar slices of mice, it was shown epitopes are probably located on the extracellular that the inward current induced by the selective ago- domains of the subunits NR2B and NR2A and are nist of the metabotropic glutamate receptors was sig- conformational. nificantly reduced in the presence of IgG that were It is interesting that later (especially with an isolated from the patients serum. After washing away increase in sample size) this hypothesis became antibodies, the current was restored. inconsistent with clinical data [39]. The NR2B subunit of NMDAR is expressed predominantly in the No less interesting are the methods used to confirm hippocampus and other areas of the forebrain. There- pathogenicity in vivo. Pioneers in this area are Coes- fore, the disruption of the function of NR1/NR2B mans et al. [58]. First, in accordance with the standard receptors will be relatively local, and is unlikely to procedure [59], eye movements in laboratory mice cause a clinical picture of the extensive neurological were analyzed. After the baseline measurements (con- deficit. The discrepancy between the putative molecu- trol data), a pump was implanted that injected purified lar mechanism and the clinical picture prompted a and concentrated IgG solution from either a patient or new hypothesis: the researchers suggested that the a healthy control into mice cerebellum (flocculus main epitopes for binding autoantibodies are located area) at a predetermined rate. All mice injected with in the much more common NR1 subunit. This patients’ IgG showed a decrease in the amplitude of hypothesis was confirmed in a number of experiments: compensatory eye movements. HEK293 cells were transfected with plasmids carrying the NR1 and/or NR2 genes. It was shown that In 2004, Pellkofer et al. [60] administered recom- patients' serum and CSF are reactive not only against binant type 1 T-helpers (Th1), which are specific for all NR1/NR2 heteromers, but also against NR1/NR1 the autologous onco-neuronal Pnma1 antigen, to homodimers, i. e. autoantibodies recognize exactly female rats. Six days after the administration of the NR1 subunit. To determine the localization of the Pnma1-specific CD4+ Th1, an inflammatory response epitope, the cells were transfected with a plasmid car- in the central nervous system was observed in rats. In rying the gene of the modified subunit NR1 (NR1d4): animals, encephalomyelitis was localized in the same it lacked amino-acid residues 25–380; however, it areas as in patients with paraneoplastic syndrome formed a complex with NR2B. Patients’ CSF and associated with these antibodies. Thus, it has been serum samples were almost non-reactive against confirmed that the reactivity of CD4+ T cells with NR1d4/NR2B complex. Thus, it was proved that the respect to onco-neuronal antigens is pathogenic and main epitope is located in the extracellular region of the autoimmune component involving T cells is one of the NR1 subunit [39]. the mechanisms of CNS damage in paraneoplastic Cell cultures also revealed that autoantibodies to syndromes. the GABAa receptor reduce the density of synaptic The same approach was used to show pathogenic- and extra-synaptic GABAaR [57]. The biochemical ity of individual antibodies: IgG from a patient with a experiment deepened the results: the researchers sug- high titer of antibodies to amphiphysin induced mus- gested that autoantibodies to GABAa enhance the internalization of receptors, which leads to a decrease cle spasms in rats (their severity correlated with the in their density. The experiment was performed as fol- dose of antibodies administered). They resemble lows: rat hippocampal neurons were incubated with spasms typical for stiff-person syndrome [61]. In 2010, the serum of patients or participants from the control in this experiment, the pathogenicity of antibodies to group, after which the surface proteins were bioti- NMDAR was confirmed [62]. nylated and isolated using granules conjugated with avi- In 2012, in vivo studies moved to a new level: the din. Neurons that were incubated with patients’ serum models of Sydenham’s chorea and PANDAS-associ- showed reduced level of surface GABAaR β3 subunits, ated disorders in rats were created by immunization whereas the total number of β3 subunits remained the same. To determine whether this effect is specific for with group A streptococcal antigen. Results from stud- GABAaR, the content of CluA1 and N-cadherin was ies [63] allows to consider this group of pathologies as evaluated in parallel and no changes were detected. encephalitis with antibodies to D2R. This experiment lead to the hypothesis of receptor The experiments described above are laborious and internalization caused by autoantibodies. technically difficult. However, the most important Neuronal cultures are also used to perform electro- results are obtained in vivo, and animal models are physiological experiments, during which it was shown imperative to reliably confirm the pathogenicity of that antibodies change potentials and currents in nerve individual autoantibodies.

NEUROCHEMICAL JOURNAL Vol. 12 No. 4 2018 AUTOIMMUNE ENCEPHALITIS 321

CONCLUSIONS: THE MODERN STATE Nowadays encephalitis with antibodies to glycine OF THE PROBLEM receptors and hyperekplexia, and a mutation in GlyR There are several important reasons for searching gene leading to hyperekplexia, a mutation in the neu- for new antibodies in AEs: rexin gene in autism and behavioral disorders similar to pathology in antineurexin antibodies encephalitis (1) Clarification of the pathogenesis of the disease, [6], involvement of the temporal lobe in AE with anti- since most of the antibodies are pathogenic, which has bodies to LGI1 and hereditary temporal epilepsy with been proven in animal experiments [64]; mutation in this gene [72] were already described. All (2) Finding antibody allows the use of immuno- these facts make AE an unique disease where it is pos- therapy and, in some cases, predicts the response to sible to evaluate the disturbances of “antibody–recep- therapy and the prognosis of the disease; tor–gene” axis functioning not only in animal models, (3) Antibody positivity may become the basis for pre- but also in different patients. scribing more effective but dangerous second line treat- Thus, in addition to the routine clinical task of ment if the refractoriness to the first-line drugs occurred. diagnosing and determination of the optimal treat- In addition, there are several problems of antibod- ment tactics, AE is a very interesting disease for further ies determination in AE: study of neurobiology of cognitive and behavioral processes, as well as immunogenetic interactions. (1) The detection and determination of all known antibodies is available only to a small number of laboratories; COMPLIANCE WITH ETHICAL STANDARDS (2) It is possible to detect low unspecific antibody Funding. No external funding was received. titers in other neurological diseases. GlyR antibodies Conflict of interest. The authors declare no conflict were found in Creutzfeldt–Jakob disease [65], of interest. NMDAR in Creutzfeldt–Jakob disease [66] and MELAS syndrome [67], GABAaR in genetically con- Ethical approval. This article does not contain any firmed Goettington’s disease [68], and sometimes studies with human participants or experimental ani- antineuronal antibodies are detected in healthy volun- mals performed by any of the authors. teers [69]; (3) Most of the available studies focus on IgG REFERENCES detection, whereas the role of IgA and IgM antibodies 1. Vitaliani, R., Mason, W., Ances, B., Zwerdling, T., is unknown, although cell cultures demonstrate Jiang, Z., and Dalmau, J., Ann. Neurol., 2005, vol. 58, changes in the presentation of the NMDA receptor pp. 594–604. under the action of these antibodies [70]; 2. Leypoldt, F., Armangue, T., and Dalmau, J., Ann. N. Y. (4) There is no correlation between the antibody Acad. Sci., 2015, vol. 1338, pp. 94–114. titer and the severity of clinical manifestations, there 3. Herken, J. and Prüss, H., Front. Psychiatry, 2017, are no clear protocols for treatment (choice of cyto- vol. 16, pp. 8–25. statics, etc.), treatment is mostly empirical. 4. Graus, F., Titulaer, M.J., Balu, R., Benseler, S., Bien, C.G., Cellucci, T., Cortese, I., Dale, R.C., Gelfand, J.M., Despite the short-term existence of this disease as a Geschwind, M., Glaser, C.A., Honnorat, J., Höft- clinical nosology, nowadays AE may be considered as berger, R., Iizuka, T., Irani, S.R., Lancaster, E., Ley- a unique disease. poldt, F., Prüss, H., Rae-Grant, A., Reindl, M., Ros- First of all, antibodies detected in AE with different enfeld, M.R., Rostásy, K., Saiz, A., Venkatesan, A., psychiatric symptoms may be one of the possible Vincent, A., Wandinger, K.P., Waters, P., and Dalmau, J., “molecular substrata” of psychiatric diseases and may Lancet Neurol., 2016, vol. 15, no. 4, pp. 391–404. 5. Davydovskaya, M.V., Boyko, A.N., Beliaeva, I.A., lead to new concepts of pathogenesis of psychiatric Martynov, M.Y., and Gusev, E.I., Zh. Nevrol. Psikhi- disorders. Using AE antibodies scientists can study atr. im. S.S. Korsakova, 2015, vol. 115, no. 4, pp. 95– cognitive processes, for example, short-term memory. 101. Secondly, it is quite interesting to consider AE from 6. Fukata, M., Yokoi, N., and Fukata, Y., Curr. Opin. the point of view of switching off the receptor func- Neurobiol., 2018, vol. 48, pp. 1–8. tion: now AE is considered together with its “genetic 7. Coyle, J.T., Adv. Neurobiol., 2017, vol. 15, pp. 255–280. twins,” diseases with a genetic mutation in the gene of 8. Pollak, T.A., McCormack, R., Peakman, M., Nichol- the receptor or another protein that is an antigen. son, T.R., and David, A.S, Psychol. Med., 2014, vol. 44, no. 12, pp. 2475–2487. As an example, the identified mutations in GRIN1, 9. Crozatti, L.L., de Brito, M.H., Lopes, B.N., and de as well as GRIN2A (NR2A) and GRIN2B (NR2B) Campos, F.P., Autops. Case. Rep., 2015, vol. 5, no. 3, lead to the development of epileptic encephalopathies. pp. 43–47. The clinical phenotype for GRIN1-encephalopathy is 10. Freitas, D.R., Santos, J.B., and Castro, C.N., Rev. Soc. described as a combination of hyperkinesis, epileptic Bras. Med. Trop., 2014, vol. 47, no. 2, pp. 260–261. seizures, and sleep-wake disorders, which is very similar 11. Yao, Y., Huang, E., Xie, B., and Cheng, Y., Neurol. to the pattern of AE with antibodies to NMDAR [71]. Sci., 2012, vol. 33, no. 1, pp. 99–102.

NEUROCHEMICAL JOURNAL Vol. 12 No. 4 2018 322 FOMINYKH et al.

12. Noguchi, H. and Moore, J.W., J. Exp. Med., 1913, tazzo, G.F., Piccolo, G., and De Camilli, P., N. Engl. vol. 17, pp. 232–238. J. Med., 1993, vol. 25, no. 328(8), pp. 546–551. 13. Vartdal, F., Vandvik, B., Michaelsen, T.E., Loe, K., 37. Honnorat, J., Antoine, J.C., Derrington, E., Aguera, M., and Norrby, E., Ann. Neurol., 1982, vol. 11, no. 1, and Belin, M.F., J. Neurol. Neurosurg. Psychiatry, 1996, pp. 35–40. vol. 61, no. 3, pp. 270–278. 14. Mitsonis, C.H., Kararizou, E., Dimopoulos, N., Tri- 38. Dalmau, J., Tuzun, E., Wu, H.Y., Masjuan, J., Rossi, J.E., antafyllou, N., Kapaki, E., Mitropoulos, P., Sfagos, K., Voloschin, A., Baehring, J.M., Shimazaki, H., Koide, R., and Vassilopoulos, D., Int. J. Neurosci., 2008, vol. 118, King, D., Mason, W., Sansing, L.H., Dichter, M.A., no. 9, pp. 1251–1257. Rosenfeld, M.R., and Lynch, D.R., Ann. Neurol., 2007, 15. Feigenson, K.A., Kusnecov, A.W., and Silverstein, S.M., vol. 61, pp. 25–36. Neurosci. Biobehav. Rev., 2014, vol. 38, pp. 72–93. 39. Dalmau, J., Gleichman, A.J., Hughes, E.G., Rossi, J.E., 16. Dale, R.C., Church, A.J., Surtees, R.A., Lees, A.J., Peng, X., Lai, M., Dessain, S.K., Rosenfeld, M.R., Adcock, J.E., Harding, B., Neville, B.G., and Giovan- Balice-Gordon, R., and Lynch, D.R., Lancet. Neurol., noni, G., Brain, 2004, vol. 127, pp. 21–33. 2008, vol. 7, pp. 1091–1098. 17. Church, A.J., Cardoso, F., Dale, R.C., Lees, A.J., 40. Coyle, J.T., Schizophr. Bul., 2012, vol. 38, no. 5, Thompson, E.J., and Giovannoni, G., Neurology, pp. 920–926. 2002, vol. 59, no. 2, pp. 227–231. 41. Irani, S.R., Alexander, S., Waters, P., Kleopa, K.A., 18. Baizabal-Carvallo, J.F. and Jankovic, J., J. Neurol. Sci., Pettingill, P., Zuliani, L., Peles, E., Buckley, C., Lang, B., 2018, vol. 385, pp. 175–184. and Vincent, A., Brain, 2010, vol. 133, no. 9, pp. 2734– 19. Brain, L., Jellinek, E.H., and Ball, K., Lancet, 1966, 2748. vol. 2, no. 7462, pp. 512–514. 42. Hutchinson, M., Waters, P., McHugh, J., Gorman, G., O’Riordan, S., Connolly, S., Hager, H., Yu, P., 20. Kishitani, T., Matsunaga, A., Ikawa, M., Hayashi, K., Becker, C.M., and Vincent. A., Neurology, 2008, Yamamura, O., Hamano, T., Watanabe, O., Tanaka, K., vol. 71, no. 16, pp. 1291–1292. Nakamoto, Y., and Yoneda, M., Medicine (Baltimore), 2017, vol. 96, no. 10, pp. e6181. 43. Lai, M., Hughes, E.G., Peng, X., Zhou, L., Gleich- man, A.J., Shu, H., Matà, S., Kremens, D., Vitaliani, R., 21. Montagna, G., Imperiali, M., Agazzi, P., D’Aurizio, F., Geschwind, M.D., Bataller, L., Kalb, R.G., Davis, R., Tozzoli, R., Feldt-Rasmussen, U., and Giovanella, L., Graus, F., Lynch, D.R., Balice-Gordon, R., and Dal- Autoimmun. Rev., 2016, vol. 15, no. 5, pp. 466–476. mau, J., Ann. Neurol., 2009, vol. 65, no. 4, pp. 424– 22. Payer, J., Petrovic, T., Lisy, L., and Langer, P., Int. J. 434. Endocrinol. Metab., 2012, vol. 10, no. 2, pp. 506–514. 44. Lancaster, E., Lai, M., Peng, X., Hughes, E., Constan- 23. Ochi, H., Horiuchi, I., Araki, N., Toda, T., Araki, T., tinescu, R., Raizer, J., Friedman, D., Skeen, M.B., and Sato, K., FEBS Lett., vol. 528, nos. 1–3, pp. 197– Grisold, W., Kimura, A., Ohta, K., Iizuka, T., Guz- 202. man, M., Graus, F., Moss, S.J., Balice-Gordon, R., 24. Yoneda, M., Fujii, A., Ito, A., Yokoyama, H., Nak- and Dalmau, J., Lancet Neurol., 2010, vol. 9, no. 1, agawa, H., and Kuriyama, M., J. Neuroimmunol., 2007, pp. 67–76. vol. 185, nos. 1–2, pp. 195–200. 45. Lancaster, E., Martinez-Hernandez, E., Titulaer, M.J., 25. Schulz, P. and Prüss, H., J. Hist. Neurosci., 2015, Boulos, M., Weaver, S., Antoine, J.C., Liebers, E., vol. 24, no. 4, pp. 371–377. Kornblum, C., Bien, C.G., Honnorat, J., Wong, S., 26. Denny-Brown, D., J. Neurol. Neurosurg. Psychiatry, Xu, J., Contractor, A., Balice-Gordon, R., and Dal- 1948, vol. 11, no. 2, pp. 73–87. mau, J., Neurology, 2011, vol. 77, no. 18, pp. 1698– 27. Guichard, M. and Vignon, G., Le Journal de Médecine 1701. de Lyon, 1949, vol. 30, pp. 197–207. 46. Carr, I., Lancet, 1982, vol. 10, no. 1(8276), pp. 844– 28. Brierley, J.B., Corsellis, J.N., Hierons, R., and Nevin. S., 845. Brain, 1960, vol. 83, pp. 357–368. 47. Boronat, A., Gelfand, J.M., Gresa-Arribas, N., 29. Corsellis, J.A., Goldberg, G.J., and Norton, A.R., Jeong, H.Y., Walsh, M., Roberts, K., Martinez-Her- Brain, 1968, vol. 91, pp. 481–496. nandez, E., Rosenfeld, M.R., Balice-Gordon, R., 30. Russell, D.S., Encephalitides, Amsterdam: Elsevier, Graus, F., Rudy, B., and Dalmau, J., Ann. Neurol., 1961, pp. 131–135. 2013, vol.73, no. 1, pp. 120–128. 31. Jaeckle, K.A., Graus, F., Houghton, A., Cardon- 48. Petit-Pedrol, M., Armangue, T., Peng, X., Bataller, L., Cardo, C., Nielsen, S.L., and Posner, J.B., Ann. Neu- Cellucci, T., Davis, R., McCracken, L., Martinez-Her- rol., 1985, vol. 18, no. 5, pp. 592–600. nandez, E., Mason, W.P., Kruer, M.C., Ritacco, D.G., Grisold, W., Meaney, B.F., Alcalá, C., Sillevis-Smitt, P., 32. Graus, F., Cordon-Cardo, C., and Posner, J.B., Neu- Titulaer, M.J., Balice-Gordon, R., Graus, F., and Dal- rology, 1985, vol. 35, no. 4, pp. 538–543. mau, J., Lancet Neurol., 2014,vol. 13, no. 3, pp. 276– 33. Graus, F., Elkon, K.B., Cordon-Cardo, C., and Pos- 286. ner, J.B., Am. J. Med., 1986, pp. 8045–8052. 49. Zhou, C., Huang, Z., Ding, L., Deel, M.E., Arain, F.M., 34. Greenlee, J.E. and Brashear, H.R., Ann. Neurol., 1983, Murray, C.R., Patel, R.S., Flanagan, C.D., and Galla- vol.14, no. 6, pp. 609–613. gher, M.J., J. Biol. Chem., 2013, vol. 288, pp. 21458– 35. Luque, F.A., Furneaux, H.M., Ferziger, R., Rosen- 21472. blum, M.K., Wray, S.H., Schold, S.C., Jr., Glantz, M.J., 50. Sabater, L., Gaig, C., Gelpi, E., Bataller, L., Lewerenz, J., Jaeckle, K.A., Biran, H., and Lesser, M., Ann. Neurol., Torres-Vega, E., Contreras, A., Giometto, B., Compta, Y., 1991, vol. 29, no. 3, pp. 241–251. Embid, C., Vilaseca, I., Iranzo, A., Santamaría, J., 36. Folli, F., Solimena, M., Cofiell, R., Austoni, M., Tal- Dalmau, J., and Graus, F., Lancet Neurol., 2014, lini, G., Fassetta, G., Bates, D., Cartlidge, N., Bot- vol. 13, no. 6, pp. 575–586.

NEUROCHEMICAL JOURNAL Vol. 12 No. 4 2018 AUTOIMMUNE ENCEPHALITIS 323

51. Escudero, D., Guasp, M., Ariño, H., Gaig, C., 66. Fujita, K., Yuasa, T., Takahashi, Y., Tanaka, K., Sako, W., Martínez-Hernández, E., Dalmau, J., and Graus, F., Koizumi, H., Iwasaki, Y., Yoshida, M., Izumi, Y., and Neurology, 2017, vol. 89, no. 14, pp. 1471–1475. Kaji, R., J. Neuroimmunol., 2012, vol. 251, nos. 1–2, 52. Gresa-Arribas, N., Planagumà, J., Petit-Pedrol, M., pp. 90–93. Kawachi, I., Katada, S., Glaser, C.A., Simabukuro, M.M., 67. Finke, C., Prüss, H., Scheel, M., Ostendorf, F., Armangué, T., Martínez-Hernández, E., Graus, F., Harms, L., Borowski, K., Wandinger, K.P., and and Dalmau, J., Neurology, 2016, vol. 86, no. 24, Ploner, C.J., J. Neurol., 2012, vol. 259, no. 3, pp. 582– pp. 2235–2242. 584. 53. Ganesh, A. and Wesley, S.F., Neurol. Clin. Pract., 2018, vol. 8, no. 1, pp. 67–73. 68. Pettingill, P., Kramer, H.B., Coebergh, J.A., Pettingill, R., 54. Dale, R.C., Merheb, V., Pillai, S., Wang, D., Cantrill, L., Maxwell, S., Nibber, A., Malaspina, A., Jacob, A., Murphy, T.K., Ben-Pazi, H., Varadkar, S., Aumann, T.D., Irani, S.R., Buckley, C., Beeson, D., Lang, B., Waters, P., Horne, M.K., Church, A.J., Fath, T., and Brilot, F., and Vincent, A., Neurology, 2015, vol. 84, no. 12, Brain, 2012, vol. 135, pp. 3453–3468. pp. 1233–1241. 55. Voltz, R., Gultekin, S.H., Rosenfeld, M.R., Gerstner, E., 69. Dahm, L., Ott, C., Steiner, J., Stepniak, B., Teegen, B., Eichen, J., Posner, J.B., and Dalmau, J., N. Engl. J. Saschenbrecker, S., Hammer, C., Borowski, K., Bege- Med.,1999, vol. 340, no. 23, pp. 1788–1795. mann, M., Lemke, S., Rentzsch, K., Probst, C., Mar- 56. Linden, D.J., Dickinson, M.H., Smeyne, M., and tens, H., Wienands, J., Spalletta, G., Weissenborn, K., Connor, J.A., Neuron, 1991, vol. 7, no. 1, pp. 81–89. Stöcker, W., and Ehrenreich, H., Ann. Neurol., 2014, 57. Ohkawa, T., Satake, S., Yokoi, N., Miyazaki, Y., vol. 76, no. 1, pp. 82–94. Ohshita, T., Sobue, G., Takashima, H., Watanabe, O., 70. Lancaster, E., Leypoldt, F., Titulaer, M.J., Honnorat, J., Fukata, Y., and Fukata, M., J. Neurosci., 2014, vol. 34, Waters, P.J., Reind, M., and Höftberger, R., Ann. Neu- no. 24, pp. 8151–8163. rol., 2015, vol. 77, no. 1, pp. 183. 58. Coesmans, M., Smitt, P.A., Linden, D.J., Shigemoto, R., 71. Lemke, J.R., Geider, K., Helbig, K.L., Heyne, H.O., Hirano, T., Yamakawa, Y., van Alphen, A.M., Luo, C., Schütz, H., Hentschel, J., Courage, C., Depienne, C., van der Geest, J.N., Kros, J.M., Gaillard, C.A., Nava, C., Heron, D., Møller, R.S., Hjalgrim, H., Lal, D., Frens, M.A., and de Zeeuw, C.I., Ann. Neurol., 2003, Neubauer, B.A., Nürnberg, P., Thiele, H., Kurlemann, G., vol. 53, no. 3, pp. 325–336. Arnold, G.L., Bhambhani, V., Bartholdi, D., Pedu- 59. van Alphen, A.M., Stahl, J.S., and De Zeeuw, C.I., rupillay, C.R., Misceo, D., Frengen, E., Strømme, P., Brain. Res., 2001, vol. 890, no. 2, pp. 296–305. Dlugos, D.J., Doherty, E.S., Bijlsma, E.K., Ruiven- 60. Pellkofer, H., Schubart, A.S., Höftberger, R., Schutze, N., kamp, C.A., Hoffer, M.J., Goldstein, A., Rajan, D.S., Pagany, M., Schüller, M., Lassmann, H., Hohlfeld, R., Narayanan, V., Ramsey, K., Belnap, N., Schrauwen, I., Voltz, R., and Linington, C., Brain, 2004, vol. 127, Richholt, R., Koeleman, B.P., Sá, J., Mendonça, C., pt. 8, pp. 1822–1830. de Kovel, C.G., Weckhuysen, S., Hardies, K., De 61. Sommer, C., Weishaupt, A., Brinkhoff, J., Biko, L., Jonghe, P., De Meirleir, L., Milh, M., Badens, C., Wessig, C., Gold, R., and Toyka, K.V., Lancet, 2005, Lebrun, M., Busa, T., Francannet, C., Piton, A., vol. 365, no. 9468, pp. 1406–1411. Riesch, E., Biskup, S., Vogt, H., Dorn, T., Helbig, I., 62. Hughes, E.G., Peng, X., Gleichman, A.J., Lai, M., Michaud, J.L., Laube, B., and Syrbe, S., Neurology, Zhou, L., Tsou, R., Parsons, T.D., Lynch, D.R., Dal- 2016, vol. 7, no. 86(23), pp. 2171–2178. mau, J., and Balice-Gordon, R.J., J. Neurosci., 2010, 72. Dazzo, E., Santulli, L., Posar, A., Fattouch, J., Conti, S., vol. 30, no. 17, pp. 5866–5875. Lodén-van Straaten, M., Mijalkovic, J., De Bortolim M., 63. Brimberg, L., Benhar, I., Mascaro-Blanco, A., Alva- Rosa, M., Millino, C., Pacchioni, B., Di Bonaventura, C., rez, K., Lotan, D., Winter, C., Klein, J., Moses, A.E., Giallonardo, A.T., Striano, S., Striano, P., Parmeg- Somnier, F.E., Leckman, J.F., Swedo, S.E., Cunning- giani, A., and Nobile, C., Epilepsy Res., 2015, vol. 110, ham, M.W., and Joel, D., Neuropsychopharmacology, pp. 132–138. 2012, vol. 37, no. 9, pp. 2076–2087. 64. Dalmau, J., Ann. Neurol., 2009, vol. 65, no. 4, pp. 424– 73. Antoine, J.C. and Honnorat J., Rev. Neurol. (Paris), 434. 2000, vol. 156, no. 1, pp. 2333. 65. Angus-Leppan, H., Rudge, P., Mead, S., Collinge, J., 74. Sabater, L., Höftberger, R., Boronat, A., Saiz, A., Dal- and Vincent, A., JAMA Neurol., 2013, vol. 70, no. 7, mau, J., and Graus, F., PLoS One, 2013, vol. 8, no. 3, pp. 919–922. p. e60438.

NEUROCHEMICAL JOURNAL Vol. 12 No. 4 2018